1,3-Bis( 3,4-dichlorophenyl)triazene from Propanil in Soils Jack R. Plimmer, Philip C. K e a r n e y , H i d e o C h i s a k a , Joseph B. Yount, and Ute I. Klingebiel
A major unknown product was isolated from 14C-ring-1abeled 3,4-dichloropropionanilide (propanil) in a Chikugo light clay soil. The identity of the unknown was established as 1,3-bis(3,4dichloropheny1)triazene by comparison of its physical and chemical properties with a synthetic sample. Infrared, ultraviolet, nuclear magnetic resonance, and mass spectra were determined. The triazene
ecently a number of condensation products arising from 3,4-dichloroaniline have been reported in soils receiving high dosages of certain acylanilide herbicides. Bartha and Pramer (1967) showed that 3,4-dichloroaniline from 3,4-dichloropropionanilide (propanil) was converted to 3,3',4,4'-tetrachloroazobenzene (TCAB) in Nixon sandy loam soil. The formation of azobenzenes from anilines depended o n the position of chlorine substituents as well as o n a number of other factors (Bartha et al., 1968). Condensation of three molecules of 3,4-dichloroaniline has now been reported by Rosen et al. (1969), and the implications of their findings have necessitated further study of the fate of aniline-based herbicides in soils. We have demonstrated that acylated anilines in soils result from microbial conversion of the starting material 3,4-dichloroaniline (Plimmer and Kearney, 1969). Katz et al. (1969) examined the possibility that diazotization reactions in rumen fluid could convert a n aniline into a reactive diazonium compound. Further reaction with another molecule of free aniline or a reactive substrate might afford a n azo compound. Their findings were negative, but d o not exclude the possibility that such reactions might occur under suitable environmental conditions. A study of 14C-ringlabeled propanil, incubated for 25 days in a number of Japanese soils (Chisaka and Kearney, 1970), provided evidence for yet another major product not previously reported in soils. The present paper describes the isolation, identification, and probable irz cico synthesis of this unknown compound in soils.
R
EXPERIMENTAL
Mass spectra were obtained with a Perkin Elmer Model GC 270 instrument or (high resolution) o n a C E C Model 21-1 10 instrument. Infrared spectra were determined as KBr disks o n Perkin Elmer Model 621 infrared spectrophotometer. Isolation of 1,3-Bis(3,4-dichlorophenyl)triazene. A Japanese light clay (Chikugo) soil treated with 14C-propaniluniformly ring labeled (0.77 pCi per SO g of soil) and sufficient purified propanil to give a concentration of 850 ppm in soil was incubated for 25 days. A sample (100 g oven-dry basis) was extracted with 200 ml of acetone, as described by Chisaka and Kearney (1970). The acetone extract was filtered and its Crops Research Division, Agricultural Research Service, U. S. Department of Agriculture, Beltsville, Md. 20705
probably arises by formation of a n intermediate diazonium cation which subsequently couples with 3,4-dichloroaniline. It is proposed that soil nitrite initially reacts with 3,4-dichloroaniline to form the intermediate diazonium cation. Consequently soil nitrite content may determine the subsequent pathways of 3,4-dichloroaniline metabolism into various condensation products or formation of the triazene.
volume reduced. The aqueous layer was extracted with benzene (three volumes of 25 ml). The benzene extract was dried (MgS04) and concentrated in a rotary evaporator. The benzene solution was applied t o one corner of a 20 crn X 20 cm thin-layer chromatographic plate coated with a 2 mm layer of silica gel GF (Merck). Benzene was used as the first developing solvent; the plate was then dried, rotated through 90°, and developed with Solvent A (hexane, benzene, acetone 7 :3 :1, v/v). Chromatograms were exposed to no-screen x-ray film for 10 days. A number of yellow spots visible on the silica plate corresponded in position to spots visible o n the developed x-ray film. The most intensely radioactive spot (Rr in benzene 0.68; R r in Solvent A 0.30) was selected for further examination. The unidentified spot was scraped from the plate and extracted from the silica gel several times with a little absolute ethanol. The combined washings were filtered through diatomaceous earth and evaporated to dryness. The pale orange-yellow residue was submitted to further analyses by mass, nuclear magnetic resonance (nmr), ultraviolet, and infrared spectrometry. Synthesis of 1,3-Bis(3,4-dichlorophenyl)triazene. The compound was synthesized by the method described in the literature for l ,3-bis(phenyl)triazene (diazoaminobenzene) (Vogel, 1956). 3,4-Dichloroaniline was used as starting material and quantities were varied accordingly. A yellow solid was collected and purified by treatment with charcoal and crystallization from petroleum ether b.p. 60-110" C , followed by recrystallization from hexane. It crystallized as pale-yellow needles, m.p. 153" C , showing a transition to rods at 145" C [Usui and Matsumura (1967) give m.p. 151" C]. The chromatographic and physical properties were identical with those of the compound isolated from soil. The molecular formula C1?H7N3Clawas confirmed by high resolution mass spectrometry, as were the compositions of the fragment ions CliHiNCll and CcH5NC1?. RESULTS AND DISCUSSION
I n the preceding paper (Chisaka and Kearney, 1970) the techniques used for extracting soils and obtaining herbicide metabolites have been described. If individual products separated by thin-layer chromatography contained a significant amount of the l'C-label, efforts were made to isolate and identify them. A compound isolated from a Japanese clay soil (Chikugo) contained more than 60% of the total extractJ. AGR. FOOD CHEM., VOL. 18, NO. 5 , 1970 859
'
I
NaOAc
Acid
Figure 1. Formation and rearrangement of 1,3-bis(3,4-dichloropheny1)triazene
~
~
E
P
mje161
mie 1 4 5
m/e 173
Figure 2a. Fragmentation of 1,3-bis(3,4-dichlorophenyl)triazene
CI
PII
m,e 188
m/e 160
CIQFiiN CI
Y
mie 173
Figure 2b. Fragmentation of a tetrachloroaminoazobenzene (Katz et al., 1969)
able radioactivity (25 of the amount initially added). O n thin-layer chromatograms, the unknown compound appeared as a yellow spot, less mobile than 3,3',4,4'-tetrachloroazobenzene. As the compound was not sufficiently volatile for gas chromatography, we suspected that it might be 4-(3,4-dichloroanilino)-3,3',4'-trichloroazobenzenemje443, reported by Rosen et a/. (1969). However, a sample examined by mass spectrometry had molecular ion at rnje 333. The ultraviolet spectrum in ethanol had maxima at 244, 305, and 363 nni, compared with 240,322, and 443 nm for 3,3',4,4'tetrachloroazobenzene, and with 238, 270, and 331 nm for 4,4'-dichloroazoxybenzene. High resolution mass spectrometry showed that the molecular ion of rnje 333 had the empirical formula CI2H7N3Cla. Fragment ions at m/e 305 and 161 had the formula CI2H7N3jCl4 and CsHsNW12, respectively.
860 J. AGR. FOOD CHEM., VOL. 18, NO. 5 , 1970
This evidence eliminates possible oxygenated structures. The presence of three nitrogen atoms is in agreement with the inference that rnje 333, an odd numbered mass, is truly the molecular ion. The formula CI2H7N3Cl4 represents several possibilities. Of these, a triazene (diazoaminobenzene) or a n aminoazobenzene appear most likely. Katz et a/. (1969) obtained isomeric aminoazobenzenes by reaction of 3,4dichloroaniline with nitrous acid. Under the conditions used, the isomeric aminoazobenzenes are probably formed in this reaction (Figure 1). Katz et al. (1969) reported the mass spectra of two compounds mje 333 which they proposed were 6-amino-2,3,3 ',4'-tetrachloroazobenzene (11) and 2amino-3 ',4,4',5-tetrachloroazobenzene (111). The molecular weight and fragmentation pattern support the aminoazobenzene structures. However, fragmentation of (I) obtained from soil revealed ions at m / e 305 (IV) by loss of nitrogen and mje 173 (V) by loss of Cl2CdH3NH (Figure 2a). A fragment at mje 161 (CsHiCl?N)(VI) must arise by a rearrangement process. In contrast, Katz et d. (1969) reported a different fragmentation pathway. Their tetrachloroaminoazobenzenes did not appear to lose nitrogen, but fragmented by loss of CsH& to give an aminodiazonium ion mje 188 (VII) (Figure 2b). The fragment rnje 173 (V) is common to both schemes. They did not observe a fragment of rnje 161 but a n ion mje 160 (VIII) was obtained, possibly by loss of nitrogen from the ion rnje 188. The preparation of diazoaminobenzene (Vogel, 1956) involves the reaction of benzenediazonium chloride with aniline in the presence of sodium acetate. Diazoaminobenzene can be isolated and is stable, but may rearrange under acidic conditions. The reaction of 3,4-dichloroaniline under analogous conditions afforded 1,3-bis(3,4-dichlorophenyl)triazene as a yellow solid. The infrared spectrum of the authentic triazene was superimposable on that of the unknown compound from soil. A broad absorption in the infrared a t 3420-3460 cm-I may be ascribed to the N H stretching vibration (Nakanishi, 1962) of a secondary amine. A second weaker band might be observed in this region if a primary aromatic amino group were present (Dolinsky and Jones, 1954). The nmr spectrum was consistent with the allocated structure. The mass spectrum, ultraviolet absorption, and chromatographic behavior of the unknown compound were identical with those of authentic 1,3-bis(3,4-dichlorophenyl)triazene. Transformations of pesticides in soils are the result of the interaction of a number of processes. Soil characteristics have been recognized as important factors in determining the quantitative relationships between the metabolites of a pesticide. On the other hand, qualitative relationships may be related to the microbial flora of the soil. The conversion of 3,4-dichloroaniline to a triazene probably requires the formation of an intermediate diazonium cation, which subsequently reacts with free 3,4-dichloroaniline. The diazonium cation may result from the reaction between dichloroaniline and soil nitrite. The formation of triazene would therefore depend on nitrite content or fertility status of the soil. In the previous paper, the detection of amounts of triazene in a soil which had previously yielded relatively large amounts of TCAB was reported. This observation suggests that if several crops of rice seedlings are returned to the soils between experiments, the nitrite level and consequently the nature of the terminal residues may be profoundly affected. Our findings therefore lead to the conclusion that the fertility status of the soil could affect the course of metabolism of certain pesticides.
LITERATURE CITED
Nakanishi, K., “Infrared Absorption Spectroscopy,” p. 38, HoldenDay, San Francisco (1962). Plimmer, J. R., Kearney, P. C., Abstr. 29, 158th Meeting ACS, New York, N. Y. (Sept. 1969). Rosen. J. D.. Siewerski. M.. Winnett. G.. Abstr. 24. 158th Meeting ACS, New York, N. Y . (Sept. 1969). Usui, Y., Matsumura, C., Yakaguku Zasshi 87,43 (1967). Vogel, A. I., “A Text-Book of Practical Organic Chemistry,” p. 626, Longmans, Green and Co., New York (1956).
Bartha, R., Linke, H. A., Pramer, D., Science 161, 582 (1968). Bartha, R., Pramer, D., Science 156, 1617 (1967). Chisaka, H., Kearney, P. C., J. AGR.FOODCHEM.(in press) (1970). Dolinsky, M., Jones, J. H., J. Ass. Ofic. Agr. Chem. 37, 197 (1954). Katz, S . E., Fassbender, C. A., Strusz, R. G . , J. Ass. Ofic. Anal. Chem. 52,1213 (1969).
Received for review April 7, 1970. Accepted July 8, 1970. Paper presented at ACS Meeting, Division of Pesticide Chemistry, Toronto, 1970. Mention of a trademark name or a proprietary product does not constitute a guarantee or warranty of the product by the USDA, and does not imply its approval to the exclusion of other products that may also be suitable.
ACKNOWLEDGMENT
Acknowledgment is made to John M. Ruth, Entomology Division, ARS, U.S. Department of Agriculture, Beltsville, Md., for determination of high resolution mass spectra.
Y
J. AGR. FOOD CHEM., VOL,. 18, NO. 5 , 1970 861
